scsi: pm80xx: Convert 'long' mdelay to msleep

For delays longer than 20ms [um]delay isn't recommended.

pm80xx_chip_soft_rst starts off with a 500ms delay before it even gets
around to checking for the results of the reset. As long as it's at least
500ms it doesn't matter what the scheduler is doing.  The delay in the
pm8001_exec_internal_task_abort does nothing, and theory is this is a delay
to avoid a double-free.

Link: https://lore.kernel.org/r/20191114100910.6153-5-deepak.ukey@microchip.comAcked-by: Jack Wang <jinpu.wang@cloud.ionos.com>
Signed-off-by: Vikram Auradkar <auradkar@google.com>
Signed-off-by: Deepak Ukey <deepak.ukey@microchip.com>
Signed-off-by: Viswas G <Viswas.G@microchip.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>

drivers/scsi/pm8001/pm80xx_hwi.c

@@ -1241,7 +1241,7 @@ pm80xx_chip_soft_rst(struct pm8001_hba_info *pm8001_ha)
 		pm8001_printk("reset register before write : 0x%x\n", regval));
 
 	pm8001_cw32(pm8001_ha, 0, SPC_REG_SOFT_RESET, SPCv_NORMAL_RESET_VALUE);
-	mdelay(500);
+	msleep(500);
 
 	regval = pm8001_cr32(pm8001_ha, 0, SPC_REG_SOFT_RESET);
 	PM8001_INIT_DBG(pm8001_ha,
@@ -2986,7 +2986,7 @@ hw_event_sas_phy_up(struct pm8001_hba_info *pm8001_ha, void *piomb)
 	pm8001_get_attached_sas_addr(phy, phy->sas_phy.attached_sas_addr);
 	spin_unlock_irqrestore(&phy->sas_phy.frame_rcvd_lock, flags);
 	if (pm8001_ha->flags == PM8001F_RUN_TIME)
-		mdelay(200);/*delay a moment to wait disk to spinup*/
+		msleep(200);/*delay a moment to wait disk to spinup*/
 	pm8001_bytes_dmaed(pm8001_ha, phy_id);
 }